Generative Adversarial Networks Based on Transformer Encoder and Convolution Block for Hyperspectral Image Classification

Nowadays, HSI classification can reach a high accuracy when given sufficient labeled samples as training set. However, the performances of existing methods decrease sharply trained on few samples. Existing in few-shot problems usually require another dataset order to improve accuracy. cross-domain problem exists these because significant spectral shift between target domain and source domain. Considering above issues, we propose new method without requiring external through combining Generative Adversarial Network, Transformer Encoder convolution block unified framework. The proposed has both global receptive field provided by local block. Experiments conducted Indian Pines, PaviaU KSC datasets demonstrate that our exceeds results deep learning for hyperspectral image problem.